Halloysite Clay Nanotubes for Enzyme Immobilization

Article pubs.acs.org/Biomac

Halloysite Clay Nanotubes for Enzyme Immobilization Joshua Tully,† Raghuvara Yendluri,† and Yuri Lvov*,†,‡ †

Institute for Micromanufacturing and Biomedical Engineering Program, Louisiana Tech University, Ruston, Louisiana, United States Ural Federal University, Ekaterinburg, Russia

‡

ABSTRACT: Halloysite clay is an aluminosilicate nanotube formed by rolling flat sheets of kaolinite clay. They have a 15 nm lumen, 50−70 nm external diameter, length of 0.5−1 μm, and different inside/outside chemistry. Due to these nanoscale properties, they are used for loading, storage, and controlled release of active chemical agents, including anticorrosions, biocides, and drugs. We studied the immobilization in halloysite of laccase, glucose oxidase, and lipase. Overall, negatively charged proteins taken above their isoelectric points were mostly loaded into the positively charged tube’s lumen. Typical tube loading with proteins was 6−7 wt % from which onethird was released in 5−10 h and the other two-thirds remained, providing enhanced biocatalysis in nanoconfined conditions. Immobilized lipase showed enhanced stability at acidic pH, and the optimum pH shifted to more alkaline pH. Immobilized laccase was more stable with respect to time, and immobilized glucose oxidase showed retention of enzymatic activity up to 70 °C, whereas the native sample was inactive.

■

INTRODUCTION

charged proteins will have an additional driving force for adsorption to the lumen of the tube.16 The fact that enzymatic catalysis occurs inside of the lumen was first shown by the urease-catalyzed hydrolysis of urea. The reaction resulted in a higher pH inside the tubes and caused the precipitation of Ca2+ and CO32− ions;16 another experiment of the same spirit utilized Fe2+ ions, forming insoluble Fe3O4.17 The synthesized calcium carbonate or magnetite filled the inner tube lumen. This technique may find applications as a means of fabricating complex core−shell type nanomaterials composed of different substances ensuing from the biomineralization reactions. This nanoconfined “enzymatic” approach may be utilized either for loading the interior of halloysite with solidified salts or with biocatalyzed organic products. Enzymes immobilized into halloysite may be “fed” with a substrate through the tube opening, realizing nanoconfined biocatalysis. Electrostatic interactions enhance the ability of the nanotubes to adsorb anionic proteins inside of the lumen. By taking proteins at a pH above their isoelectric point, one institutes a negative net charge and efficiently coerces the proteins into positively charged lumen. Thermogravimetric analysis can determine the mass of the loaded proteins and UV−vis spectroscopy gives desorption kinetics and the amount of proteins available for release.14 We studied immobilized laccase, glucose oxidase, lipase, and pepsin into halloysite clay nanotubes. For negatively charged proteins, halloysite could support 5−7 wt % loading, but only about one-third of these proteins were released within 5−10 h, while the remaining proteins stayed inside and were functional

Enzymes are among the most powerful catalysts. However, enzymes can only be used in highly controlled aqueous environments with mild conditions. There are several methods that one can use to overcome this limitation. Point mutations of the constituent amino acids can enhance the enzymes stability.1 Cross-linking is simpler but often significantly reduces the activity of the enzyme.2,3 Porous substrate immobilization is another widely used method because not only does the technique enhance the stability of the enzyme, it also promotes reuse if the adsorbent allows. Immobilization is simple and cheap, but it often comes at the cost of catalytic efficiency. There are many types of substrates were used for immobilization including zeolites,4,5 natural and synthetic polymers, and various nano/micro particles.6−8 Dispersions of platy montmorillonite clay were used for immobilization of invertase, cellulase, peroxidase, and laccase.9−11 Halloysite, an aluminosilicate nanotube is a particularly interesting substrate because of its 15 nm lumen that allows for the immobilization inside of the tube of various molecules including drugs, antiseptics, anticorrosion agents and proteins.12−15 Halloysite nanotubes are a cheap, widely available nanoclay, and interest in this material outside of the ceramic industry is rapidly rising.14 The length of the nanotubes ranges between 400 and 1000 nm, the outer diameter is between 50 and 70 nm, and the lumen diameter is 15−20 nm. The distinguishing feature of halloysite is the positively charged inner lumen, which consists mostly of aluminum hydroxide, and the negatively charged surface, which is silicon dioxide.14 At moderate pH (between 3 and 10) the tubes’ inside/outside surfaces maintain opposing charges and enable selective immobilization of charged molecules, such that negatively © 2015 American Chemical Society

Received: November 13, 2015 Revised: December 21, 2015 Published: December 23, 2015 615

DOI: 10.1021/acs.biomac.5b01542 Biomacromolecules 2016, 17, 615−621

Article

Biomacromolecules

on a Thermal Advantage Q50. UV−visible results were obtained using a spectrophotometer by Agilent P45. Immobilization of Proteins Using Halloysite. The chosen proteins were mixed at a 2:1, halloysite to enzyme, ratio. Typical volumes are 1 mL for samples containing less than 200 mg of halloysite and up to 10 mL for samples above 1 g of nanotubes. After the addition of water, the sample is vortexed and sonicated until there are no visible aggregates of halloysite. The sample is then placed into a vacuum and the pressure is reduced to approximately 200 mmHg for 15 min. Then pressure is allowed to come to ambient conditions, the sample is agitated, and the process is repeated for a total of three vacuum cycles, allowing the substitution of inner air with enzyme solutions. After vacuuming, the sample is centrifuged at 7000 rpm for 3 min. The supernatant can be discarded or measured and the samples are washed with water twice. The sample is then dried overnight in a desiccator, ground, and used for the rest of the study. Enzymatic Activity of Lipase. The enzymatic activity of lipase was studied using a modified method.18 In short, 2.3 mL of 50 mM Trizma buffer pH 8.3 at 25 °C with 2% Triton X-100 added to an appropriate container with 100 μL of 18.96 mM 4-nitrophenyl palmitate in 2-propanol. This solution is incubated at 37 °C for 5 min, then 100 μL of enzyme is added to the solution. The reaction is carried out for 5 min in a 37 °C water bath. After time has expired, the container is transferred to an ice bath for 3 min to inhibit the reaction. Next, 1 mL of the reaction solution is transferred to a centrifuge tube and centrifuged at 7000 rpm for 3 min at 4 °C to precipitate halloysite. The centrifuged reaction solution is then stored on ice and the absorbance of the solution is measured at 397 nm, against a blank of the buffer solution. A molar extinction coefficient of 18500 M−1 cm−1

for a longer time as compared with free enzymes in solution (Scheme 1). The temperature stability of glucose oxidase was increased up to 70 °C, and an extension of storage time for the halloysite-immobilized enzymes was observed. Scheme 1. Schematic Representation of the Enzyme Immobilization

■

MATERIALS AND METHODS

Halloysite nanotubes were supplied by Applied Minerals Inc., NY. Glucose oxidase, lipase, laccase as well as other chemicals were purchased from Sigma-Aldrich and used without further processing. Transmission electron microscopy was performed on Tecnai G2 F30 Twin, U.S.A., at 200 kV. Thermogravimetric analysis was performed

Figure 1. (a, b) TGA curves and their derivatives for pristine proteins and halloysite nanotubes; (c, d) TGA curves for the protein immobilized samples (heating rate 10 °C/min). 616

DOI: 10.1021/acs.biomac.5b01542 Biomacromolecules 2016, 17, 615−621

Article

Biomacromolecules was determined in the laboratory for 4-nitrophenol, and this value was used to determine the concentration of 4-nitrophenol in the solution. Enzymatic Activity of Laccase. The assay used to evaluate the activity of laccase was a modified method of the oxidation of ABTS.19 A total of 100 μL of appropriate dilute enzyme (50 μg/mL native or 20−40 mg/mL immobilized) is added to 1 mL of 1 mg/mL ABTS in 100 mM pH 3 citrate buffer. The reaction is incubated at room temperature for 2 min, and 10 μL of 0.1% sodium azide is added to stop the reaction. The samples are centrifuged at 7000 rpm for 3 min, and the absorbance of the solution is measured at 420 nm. A molar extinction coefficient for oxidized ABTS was determined to be 19589 M−1 cm−1, and this value was used to calculate the concentration of oxidized ABTS in solution. Enzymatic Activity of Glucose Oxidase. We used a method based on a procedure using 10-acetyl-3,7-dihydroxyphenoxazine.20 A total of 50 μL of a working solution containing 100 μM 10-acetyl-3,7dihydroxyphenoxazine, 0.2 units/mL horse radish peroxidase, and 100 mM D-glucose is added to a 96-well plate. A 50 μL aliquot of appropriately dilute enzyme (0.1 U/mL or 0.25 mg/mL immobilized enzyme) is added to the selected wells. The reaction is measured in a fluorescent plate reader with an excitation wavelength of 540 nm and the fluorescence is read at 600 nm for 10 min, with a reading being taken every minute. The rate of the reaction is taken as the linear regression of the five most linear data points. Measuring Desorption. A dried sample of immobilized enzyme is added to 1 mL of DI H2O and a timer is begun. After the desired amount of time has passed, in this study, common intervals were 5, 10, 15, 30, or 60 min, the timer is stopped, and the sample is centrifuged at 7000 rpm for 3 min. The supernatant of the sample is removed and stored at room temperature. The sample is dispersed again, and the timer is started again. The process is repeated for the desired duration of the study at the chosen intervals. After all supernatant samples are collected, they are centrifuged at 7000 rpm for 3 min. The absorbance of the supernatant samples is measured at the characteristic wavelength of the protein being measured (240−280 nm), and for each protein, the extinction coefficient was determined. Generating Adsorption Isotherms. Adsorption isotherms were generated by mixing halloysite and the chosen enzyme in varying ratios (1:2−2:1 halloysite to enzyme). The samples are left overnight and are then centrifuged and dried in the same manner as with immobilization. Thermogravimetric analysis of the samples was performed from room temperature to 600 °C. The adsorbed mass is calculated as the percent mass loss over 200−400 °C minus the mass loss of a blank sample (one that contained only halloysite) over the same range. Measuring the Effects of pH and Temperature. In order to determine the effect of pH and temperature on the samples, the following procedure was used. For pH stability, the enzyme was diluted to the appropriate concentration into the chosen buffer. The buffers used were 100 mM citrate for pH 3−6, 100 mM phosphate buffer for pH 7, 100 mM Trizma for pH 8 and 100 mM carbonate buffer for pH 10. After incubating for 15 min, the enzyme’s residual activity was measured using the relevant method presented in the Materials and Methods. For temperature stability, the enzyme is diluted to the proper concentration in DI H2O. The enzyme solutions are incubated at the desired temperature for 15 min, cooled to room temperature, and the residual activity is measured using the appropriate technique for the enzyme.

acknowledged that organics degrade at this temperature range, and the mass lost over this range does not overlap with the decomposition of the hydroxyl groups of halloysite in the range of 480−550 °C (Figure 1d). Restricting the mass loss to 200− 400 °C may underestimate the mass of protein in the sample. One can observe a small maximum degradation temperature shift for laccase and lipase and immobilized glucose oxidase responds with a larger shift. This shift may be due to stabilization of the hydroxyl molecules by proteins adsorbed onto the surface of the lumen. TGA analysis demonstrates that proteins adsorb onto halloysite but there is an obvious skepticism of where exactly the protein may adsorb. There is no doubt that the proteins exist on the surface of the tubes, as it is the most available surface for adsorption. Using fluorescent tags one can readily identify proteins on the surface of the tubes with fluorescent microscopy. However, identifying proteins inside of the tubes is significantly more difficult. Transmission electron microscopy is helpful, but proteins have a low electron density and little contrast in comparison with the denser aluminosilicate tubes.21,22 In an attempt to overcome this issue, we added a heavy metal counterstain to aid in visualization. The aqueous heavy metal stain is added to a sample of dried halloysite on the grid. The stain is allowed to dry before images are acquired. Figure 2

Figure 2. TEM image of halloysite−lipase counterstained with uranyl acetate.

shows TEM images of halloysite tubes loaded with lipase and stained with uranyl acetate. It is apparent from the small irregular formations inside of the lumen that there is stained protein inside of the tube. In Figure 3 we show results for a standard approach visualization of immobilized proteins. Using the fluorescent probe 8-anilinonaphthalene-1-sulfonic acid (ANS), we were able to see proteins on the tube aggregates. The probe undergoes a blue shift when introduced to a hydrophobic environment. Therefore, the fluorescent intensity of the tubes is significantly higher when proteins are adsorbed than in the control with simply the nanotubes and dye. Other proteins were also imaged (but not presented here) and fluoresced significantly. Therefore, it is apparent that the used enzymes were adsorbed on the clay nanotubes.

■

RESULTS AND DISCUSSION Adsorption of Proteins on Halloysite. For quantifying the mass of adsorbed protein, we used thermogravimetric analysis (TGA). This method provides greater accuracy compared to measuring the residual protein in solution after the adsorption process is performed. The mass of protein in a sample is taken as the mass percent lost over the temperature range 250−400 °C minus the mass loss of a blank sample of halloysite (∼1.3 wt %), as this is the typical degradation temperature range for proteins as seen in Figure 1b. It is widely 617

DOI: 10.1021/acs.biomac.5b01542 Biomacromolecules 2016, 17, 615−621

Article

Biomacromolecules

Figure 3. Optical and fluorescent microscopy of halloysite nanotube aggregates exposed to the fluorescent probe 8-anilinonaphthalene-1-sulfonic acid. Fluorescent intensity indicates a hydrophobic environment. The control samples contain halloysite and ANS, while the sample label BSA contained bovine serum albumin immobilized on halloysite.

Figure 4. (a) TGA curves of a lipase−halloysite composite before and after incubation overnight in water. The sample was loaded at a ratio of 1:2 protein to halloysite. The heating rate was 10 °C/min. (b) The mass of protein immobilized on halloysite in various ratios before and after release. Values are extracted from TGA curves as the mass percent lost over 200−400 °C.

Our assumption is that proteins undergo irreversible tube adsorption which would allow for sustained functional enzymatic nanoreactors. It is acknowledged that proteins, more so than other molecules, show hysteresis in adsorption processes23−26 tending to stick more or less permanently to the surfaces on which they adsorb. Figure 4a shows two TGA curves of halloysite-immobilized lipase before and after incubation overnight in water. We can see that a relatively small portion (1/3) of the protein desorbs overnight, while the other 2/3 remains. This data is in accordance with data obtained by observing the desorption of laccase and glucose oxidase from halloysite. Regardless of the sample preparation, we note that the halloysite−protein composites have two fractions of proteins: proteins that are resistant to desorption and a smaller fraction that is desorbed during few hours. The larger the initial concentration, the larger the adsorbed mass and mass that desorbs from the tubes (Figure 4b). This phenomenon is known as relaxation, that is, that proteins tend to maximize contact with a surface given enough time and a low enough concentration. Time is important for relaxation to occur as well

as concentration, increasing the interaction of the protein with the surface. As the concentration of protein goes down, one sees less desorption, indicating the proteins are “stickier” probably having enough space to accommodate on the surface. One would assume that there is a maximum amount of adsorption without multilayer formation. Figure 5 shows an

Figure 5. Adsorption isotherm for lipase, pH 6.5. The mass percent of protein of each sample was determined by TGA at a heating rate of 10 °C/min. 618

DOI: 10.1021/acs.biomac.5b01542 Biomacromolecules 2016, 17, 615−621

Article

Biomacromolecules

Proteins Loose Fraction Release Kinetics. Much of the other research done with halloysite emphasizes the ability of halloysite to slowly release molecules over time.27−29 Drugs, antibiotics, and anticorrosion molecules demonstrated 5−10 h extended release, and the fraction of halloysite’s proteins identified above as loosely bound and “releasable” as well desorb in a few hours. Glucose oxidase, laccase, and pepsin show sustained release from halloysite nanotubes, as seen in Figure 7. This released protein mass is the above-mentioned

adsorption isotherm of lipase and illustrates that the kinetics of adsorption can be describe as linear (R2 = 0.95) or fit to the Langmuir adsorption. Using nonlinear regression, the maximum mass in weight percent, of lipase that can be adsorbed onto halloysite is 10.9 ± 0.5 wt % and the model gives a Langmuir equilibrium with a constant of 0.61. The net charge of a protein is positive when the pH of the solution is below its isoelectric point and negative when the pH is above it. Isoelectric points for proteins used in this study are shown in Table 1. We demonstrate that adsorption depends on Table 1. Proteins Studied, Their Isoelectric Points, and Sizes protein albumin lipase laccase glucose oxidase pepsin

abbreviation

isoelectric point

BSA Lip Lacc GOX

pH pH pH pH

7.5−8 4.5−5.5 3.5−5 4−4.5

Peps

pH 1

molecular mass

minimal diameter/size, nm

68000 58600 55800 65600

∼5.4 ∼5.1 ∼5.0 ∼5.4

41300

∼4.4

Figure 7. Release curves for the proteins in deionized water at pH 6.5. Glucose oxidase (a) was adsorbed at pH 2 and laccase and pepsin (b, c) at pH 6.5. Curves are given by nonlinear fitting of the KorsmeyerPeppas model (c(t) = atr), with parameters a = 0.69 and r = 0.11 for glucose oxidase, a = 0.56 and r = 0.18 for laccase, and a = 0.71 and r = 0.10 for pepsin.

the pH of the solution and consequently the net charge of the proteins (Figure 6). Negatively charged proteins were

about 1/3 of total protein load into halloysite. The proteins undergo a 4−5 h burst phase in which 70−80 wt % of protein is released, followed by prolonged release over 20−30 h. Laccase and pepsin would be negative, while glucose oxidase is positive at the chosen pH; however, according to the power (Korsmeyer-Peppas model30), there does not seem to be a correlation between charge and release kinetics. Retention of Enzymatic Activity. Immobilized enzymes often have greater functional stability in pH and temperatures that would typically deactivate the native enzyme. An enzymes’ ability to catalyze a reaction depends on preservation of its conformation. When adsorbed, assuming that adsorption does not denature the protein, the protein has significantly less conformational freedom since it is anchored to the surface. Now when the enzyme is exposed to an extreme environment it is less likely to morph into an inactive conformation. We saw an increase in thermal stability for both glucose oxidase and lipase albeit the effects of immobilization manifest in different patterns. Figure 8a shows the mean relative rates of heat-treated glucose oxidase samples at four different temperatures. The mean relative rate is calculated by measuring the rate of the catalyzed reaction of a sample of glucose oxidase at room temperature then dividing the rate of the heat-treated sample against the room temperature rate. We see that at normal and moderately high temperatures that immobilization (bars labeled HNT) has a small negative impact. However, as the temperature continues to increase above 60 °C, immobilized samples retain activity, while native samples quickly fall off. In Figure 8b, one sees the stabilization effect of immobilization of lipase as compared to glucose oxidase. At moderately high temperatures, 50 and 60 °C, immobilization helps to retain an essentially enhanced enzymatic activity, but this effect is less profound as the temperature continues to increase from 50 to 70 °C. This difference in activity stabilization may be related with different adsorption mechanisms: strongly anionic glucose oxidase is adsorbed

Figure 6. Mass of immobilized proteins depending on the charge of the enzyme.

immobilized at pH 10 carbonate buffer (above their isoelectric points) and positively charged proteins were immobilized at pH 2 in phosphate buffer (below their isoelectric points). The bottom label of Figure 6 indicates the protein immobilized and the legend indicates the net charge of the protein in solution. We see that the proteins adsorb in much higher mass when they are positively charged than when they are negatively charged. This is likely because the outer surface of halloysite is negatively charged above pH 2, while most proteins are positively charged, which increases their interaction with the surface of the tube and minimized interaction with the lumen. At alkaline pH, the outer surface of the tube is strongly negative and the proteins are also strongly negative, which would coerce them to adsorb in smaller amounts and mostly inside of the lumen of halloysite. Negatively charged proteins account for 2− 4 wt % and positively charged proteins, assumingly adsorbed at the larger outside surface, account for about 7 wt %. One can calculate that a monolayer of protein covering the tube’s lumen corresponds to 2−3 wt %, which is close to experimental data, while an external protein monolayer coating gives about 15 wt % of proteins to the clay nanotubes (assuming 0.5 packing coefficient for hard protein spheres with diameters taken from Table 1). 619

DOI: 10.1021/acs.biomac.5b01542 Biomacromolecules 2016, 17, 615−621

Article

Biomacromolecules

Figure 8. (a) Relative reaction rates of treated glucose oxidase for 50−70 °C. The labels indicate if the enzyme was immobilized on halloysite (HNT) or pristine (native). (b) Relative reaction rates of treated lipase within 50−70 °C.

halloysite−double hydroxide composites.24−26,32 These threeor four-component composites are complicated, but allowed, for higher enzymatic activity and may be practically useful; however, the location of the enzymes was unclear, and the relationship between the sorption process and its effects on the enzymes was not fully developed. In these works, halloysite lacked detailed analysis on the interaction between the clay nanotubes and enzyme. We tried to delineate enzyme-halloysite locations, binding rate, and sustained released fractions of the proteins, as well as how direct, nonmodified, clay nanotubes affect enzymatic stability. Finally, nanotubes coated with proteins take on a hydrophobic character, as was seen by fluorescent spectroscopy. Therefore, this method may also be used to increase the hydrophobicity of the tubes similarly to the method by Cavallaro et al.33

predominately into negatively charged lumen and cationic lipase is located mostly on the negative tubes surfaces. Besides, it is known that interface positioning is increasing catalytic activity of lipase.31 In addition to temperature stability, we demonstrated stabilization of enzymes over both time and pH. We found the latter was true for lipase, as seen in Figure 9a, but was not

■

CONCLUSIONS Halloysite nanotubes can adsorb proteins at 4−8 wt % in the lumen and on the surface of the tubes depending on the net charge of the protein. One third of the adsorbed protein will release over 10−30 h, with the majority releasing within the first 4 h. The remaining two-thirds of the protein stays on the surfaces of the tube lumen, perhaps desorbing at a miniscule rate, thus, allowing for usage of this system as enzymatic nanoreactor. Adsorption depends on the nature of the protein and the conditions in which the adsorption procedure is performed, with the main feature being electrostatic interactions between the protein’s net charge and the inner negative and outer positive charge on the nanotube. Positively charged proteins adsorb in higher amounts than negatively charged proteins due to the larger area of the negatively charged surface of the tube. Negatively charged proteins are adsorbed mostly into positively charged tube lumen, and the total adsorbed mass is halved. Adsorbed proteins show increased thermal stability and temporal biocatalytic abilities, especially at a pH above or below the enzyme’s optimum.

Figure 9. (a) Relative activity of immobilized and native lipase as a function of pH. (b) Temporal stability of laccase in both free and immobilized form. The samples were dissolved in pH 5 buffer, and a small volume was taken from the sample and assayed daily.

obvious for laccase. In this experiment, samples of lipase were incubated in a buffer of the prescribed pH and then assayed according to the Materials and Methods. Immobilized lipase fared much better than the native sample (its activity has pH maximum of pH 7.4), and it appears the optimum pH of the enzyme has been translated to higher pH. Laccase showed an essential increase of stability over time (Figure 9b), and after 8 days it was at the same level as the first day for unmodified enzyme. In some publications on enzyme immobilization on halloysite nanotubes, it was demonstrated that having additional intermediate molecules that bonds to halloysite on one end and to the enzyme on the other gives functional biocomposites. These include peroxidase and laccase immobilized to halloysite micelles with dopamine or lysozyme immobilized on

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. 620

DOI: 10.1021/acs.biomac.5b01542 Biomacromolecules 2016, 17, 615−621

Article

Biomacromolecules Notes

(31) Palomo, J. M.; Munoz, G.; Fernández-Lorente, G.; Mateo, C.; Fernández-Lafuente, R.; Guisán, J. J. Mol. Catal. B: Enzym. 2002, 19, 279−286. (32) Wang, Y.; Liu, C.; Zhang, Y.; Zhang, B.; Liu, J. ACS Sustainable Chem. Eng. 2015, 3, 1183−1189. (33) Cavallaro, G.; Lazzara, G.; Milioto, S.; Parisi, F. Langmuir 2015, 31 (27), 7472−747.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This material is based on work funded by the U.S. Department of Energy under EPSCoR Grant No. DE-SC0012432, with additional support from the Louisiana Board of Regents. Y.L. also thanks Russian Science Foundation, Grant 15-12-20021. J.T. developed this work under the GRO Fellowship Assistance Agreement No. 9177240 U.S Environmental Protection Agency (EPA). The EPA does not review this work, nor does it represent the views of the EPA.

■

REFERENCES

(1) Shoichet, B. K.; Baase, W. A.; Kuroki, R.; Matthews, B. W. Proc. Natl. Acad. Sci. U. S. A. 1995, 92 (2), 452−456. (2) Migneault, I.; Dartiguenave, C.; Bertrand, M. J.; Waldron, K. C. BioTechniques 2004, 37 (5), 790−806. (3) Cao, L.; van Rantwijk, F.; Sheldon, R. A. Org. Lett. 2000, 2 (10), 1361−1364. (4) Tavolaro, A.; Tavolaro, P.; Drioli, E. Colloids Surf., B 2007, 55 (1), 67−76. (5) Yagiz, F.; Kazan, D.; Akin, A. N. Chem. Eng. J. (Amsterdam, Neth.) 2007, 134 (1), 262−267. (6) Absolom, D. R.; Zingg, W.; Neumann, A. W. J. Biomed. Mater. Res. 1987, 21 (2), 161−171. (7) Shirahama, H.; Lyklema, J.; Norde, W. J. Colloid Interface Sci. 1990, 139 (1), 177−187. (8) Ostuni, E.; Grzybowski, B. A.; Mrksich, M.; Roberts, C. S.; Whitesides, G. M. Langmuir 2003, 19 (5), 1861−1872. (9) Lundqvist, M.; Sethson, I.; Jonsson, B. H. Langmuir 2004, 20 (24), 10639−10647. (10) Safari, S. A.; Emtiazi, G.; Shariatmadari, H. J. Colloid Interface Sci. 2005, 290, 39−44. (11) Sanjay, G.; Sugunan, S. Food Chem. 2006, 94, 573−579. (12) Joussein, E.; Petit, S.; Churchman, J.; Theng, B.; Righi, D.; Delvaux, B. Clay Miner. 2005, 40, 383−426. (13) Du, M.; Guo, B.; Jia, D. Polym. Int. 2010, 59, 574−588. (14) Lvov, Y.; Abdullayev, E. Prog. Polym. Sci. 2013, 38, 1690−1719. (15) Liu, M.; Jia, Z.; Jia, D.; Zhou, C. Prog. Polym. Sci. 2014, 39, 1498−1530. (16) Shchukin, D.; Price, R.; Lvov, Y. Small 2005, 1, 510−515. (17) Zheng, P.; Du, Y.; Ma, X. Mater. Chem. Phys. 2015, 151, 14−17. (18) Gupta, N.; Rathi, P.; Gupta, R. Anal. Biochem. 2002, 311 (1), 98−99. (19) Han, M. J.; Choi, H. T.; Song, H. G. J. Microbiol. (Seoul, Repub. Korea) 2005, 43 (6), 555−60. (20) Zhou, M.; Diwu, Z.; Panchuk-Voloshina, N.; Haugland, R. P. Anal. Biochem. 1997, 253 (2), 162−168. (21) Erickson, H. P. Biol. Proced. Online 2009, 11, 32−51. (22) Consortium, T. U. Nucleic Acids Res. 2015, 43, D204−D212. (23) Zhai, R.; Zhang, B.; Liu, L.; Xie, Y.; Zhang, H.; Liu, J. Catal. Commun. 2010, 12, 259−266. (24) Chao, C.; Liu, J.; Wang, J.; Zhang, Y.; Zhang, B.; Zhang, Y.; Xiang, X.; Chen, R. ACS Appl. Mater. Interfaces 2013, 5, 10559−1566. (25) Chao, C.; Zhang, B.; Zhai, R.; Xiang, X.; Liu, J.; Chen, R. ACS Sustainable Chem. Eng. 2014, 2, 396−403. (26) Zhai, R.; Zhang, B.; Wan, Y.; Li, C.; Wang, J.; Liu, J. Chem. Eng. J. (Amsterdam, Neth.) 2013, 214, 304−309. (27) Vergaro, V.; Lvov, Y.; Leporatti, S. Macromol. Biosci. 2012, 12, 1265−1271. (28) Lvov, Y.; Aerov, A.; Fakhrullin, R. Adv. Colloid Interface Sci. 2014, 207, 189−198. (29) Levis, S.; Deasy, P. Int. J. Pharm. (Amsterdam, Neth.) 2003, 253, 145−57. (30) Peppas, N. A. Pharm. Acta Helv. 1985, 60, 110−118. 621

DOI: 10.1021/acs.biomac.5b01542 Biomacromolecules 2016, 17, 615−621